Photometric and Spectroscopic Observations, and Abundance

Home , SN 2011fe

Photometric and spectroscopic observations, and abundance tomography modelling of the Type Ia supernova SN 2014J located in M82 Ashall, C., Mazzali, P., Bersier, D., Hachinger, S., Phillips, M., Percival, S., James, P., & Maguire, K. (2014). Photometric and spectroscopic observations, and abundance tomography modelling of the Type Ia supernova SN 2014J located in M82. Monthly Notices of the Royal Astronomical Society, 445(4), 4427-4434. https://doi.org/10.1093/mnras/stu1995 Published in: Monthly Notices of the Royal Astronomical Society

Document Version: Publisher's PDF, also known as Version of record

Queen's University Belfast - Research Portal: Link to publication record in Queen's University Belfast Research Portal

Publisher rights Copyright 2014 the authors. This is an open access article published under a Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution and reproduction in any medium, provided the author and source are cited.

General rights Copyright for the publications made accessible via the Queen's University Belfast Research Portal is retained by the author(s) and / or other copyright owners and it is a condition of accessing these publications that users recognise and abide by the legal requirements associated with these rights.

Take down policy The Research Portal is Queen's institutional repository that provides access to Queen's research output. Every effort has been made to ensure that content in the Research Portal does not infringe any person's rights, or applicable UK laws. If you discover content in the Research Portal that you believe breaches copyright or violates any law, please contact [email protected].

Download date:05. Oct. 2021 MNRAS 445, 4424–4434 (2014) doi:10.1093/mnras/stu1995

Photometric and spectroscopic observations, and abundance tomography modelling of the Type Ia supernova SN 2014J located in M82

C. Ashall,1‹ P. Mazzali,1,2,3 D. Bersier,1 S. Hachinger,4,5 M. Phillips,6 S. Percival,1 Downloaded from https://academic.oup.com/mnras/article-abstract/445/4/4424/1078876 by Queen's University of Belfast user on 16 November 2018 P. James1 and K. Maguire7 1Astrophysics Research Institute, Liverpool John Moores University, IC2, Liverpool Science Park, 146 Brownlow Hill, Liverpool L3 5RF, UK 2Istituto Nazionale di Astroﬁsica-OAPd, vicolo dell’Osservatorio 5, I-35122 Padova, Italy 3Max-Planck-Institut fur¨ Astrophysik, Karl-Schwarzschild-Str. 1, D-85748 Garching, Germany 4Institut fur¨ Theoretische Physik und Astrophysik, Universitat¨ Wurzburg,¨ Emil-Fischer-Str. 31, D-97074 Wurzburg,¨ Germany 5Institut fur¨ Mathematik, Universitat¨ Wurzburg,¨ Emil-Fischer-Str. 30, D-97074 Wurzburg,¨ Germany 6Carnegie Observatories, Las Campanas Observatory, Casilla 601, La Serena, Chile 7European Southern Observatory, D-85748 Garching bei Munchen,¨ Germany

Accepted 2014 September 22. Received 2014 September 8; in original form 2014 July 29

ABSTRACT Spectroscopic and photometric observations of the nearby Type Ia Supernova (SN Ia) SN 2014J are presented. Spectroscopic observations were taken −8to+10 d relative to B- band maximum, using FRODOSpec, a multipurpose integral-ﬁeld unit spectrograph. The observations range from 3900 to 9000 Å. SN 2014J is located in M82 which makes it the closest SN Ia studied in at least the last 28 yr. It is a spectroscopically normal SN Ia with high-velocity features. We model the spectra of SN 2014J with a Monte Carlo radiative transfer code, using the abundance tomography technique. SN 2014J is highly reddened, with a host galaxy extinction of E(B − V) = 1.2 (RV = 1.38). It has a m15(B)of1.08± 0.03 when corrected for extinction. As SN 2014J is a normal SN Ia, the density structure of the classical W7 model was selected. The model and photometric luminosities are both consistent with B-band maximum occurring on JD 245 6690.4 ± 0.12. The abundance of the SN 2014J behaves like other normal SN Ia, with signiﬁcant amounts of silicon (12 per cent by mass) and sulphur (9 per cent by mass) at high velocities (12 300 km s−1) and the low-velocity ejecta (v<6500 km s−1) consists almost entirely of 56Ni. Key words: radiative transfer – techniques: spectroscopic – supernovae: general – supernovae: individual: SN 2014J.

(WD) which accretes mass from a non-electron-degenerate com- 1 INTRODUCTION panion star (Nomoto, Iwamoto & Kishimoto 1997). In this single Supernovae are important and much-studied astrophysical events. degenerate (SD) scenario (Hoyle & Fowler 1960), the WD can ex- For example, they are the main producers of heavy elements in the plode when it approaches the Chandrasekhar mass. There are several Universe, and Type Ia supernovae (SNe Ia) produce most of the suggested ways in which this can occur, including a subsonic ex- iron-group materials (Iwamoto 1999).SNeIahavealsobeencon- plosion (a deflagration) and a supersonic explosion (a detonation) firmed as the best cosmological ‘standard candles’ and as a result as well as an explosion with a transition to a detonation. In the there has been a dramatic increase in the rate at which they are fully subsonic explosion there is not enough energy to fully power observed. They were integral in the discovery of the acceleration the SN Ia and in supersonic explosion there is too much 56Ni in of the universe (Riess 1998; Perlmutter et al. 1999), and are now the ejecta. Therefore, the transition may be the correct explosion an important cosmological probe in improving the understanding model for SN Ia (Khokhlov 1991). Two further SD explosion mod- of the nature of the positive cosmological constant. However, the els are fast deflagration and sub-Chandrasekhar-mass explosions true intrinsic properties of SN Ia are not yet fully understood, in- (Nomoto, Thielemann & Yokoi 1984; Livne & Arnett 1995). The cluding their progenitor system and diversity in luminosity (e.g. SN other suggested progenitor model is a double degenerate (DD) sce- 1991bg; Leibundgut et al. 1993). There are currently two favoured nario, where the SN results from the merger of two WDs (Iben & progenitor scenarios. The first is a carbon/oxygen White Dwarf Tutukov 1984). Thanks to the dramatic increase in observations of SN Ia, in the last two decades, it is now possible to obtain good spectral time E-mail: [email protected] series of them. These time series can span from before B-band

C 2014 The Authors Published by Oxford University Press on behalf of the Royal Astronomical Society SN 2014J models 4425 maximum to the nebular phase. One approach to fully understand Table 1. Log of spectroscopic observations of SN 2014J. the composition and progenitor system of an individual SN Ia, is to use Monte Carlo (MC) radiative transfer code and the abundance Epoch MJDa Phaseb Exp. time Instrument tomography technique (Mazzali 2000; Stehle et al. 2005). This (d) (s) approach models early-time observed spectra, by changing input 2014-01-26 56684.12 −8 500 FRODOSpec parameters such as the chemical abundance, bolometric luminos- 2014-01-27 56685.12 −7 500 FRODOSpec ity, photospheric velocity and time since explosion. The abundance 2014-02-03 56692.01 +0 600 FRODOSpec Downloaded from https://academic.oup.com/mnras/article-abstract/445/4/4424/1078876 by Queen's University of Belfast user on 16 November 2018 tomography approach exploits the fact that with time deeper and 2014-02-04 56692.90 +1 600 FRODOSpec deeper layers of the ejecta become visible. By modelling time series, 2014-02-04 56693.13 +13× 120 INT spectral information about the abundances at different depths can be 2014-02-05 56693.89 +2 600 FRODOSpec extracted, with this it is possible to reconstruct the abundance strat- 2014-02-06 56694.96 +3 600 FRODOSpec ification from the observational data. This approach directly links 2014-02-07 56695.89 +4 500 FRODOSpec + × the theoretical models and observed spectra to help one get a true 2014-02-07 56696.13 43120 INT + understanding of the early-time evolution of a SN Ia. The MC ra- 2014-02-08 56696.92 5 500 FROSOSpec 2014-02-09 56697.97 +6 500 FRODOSpec diative transfer code has been successfully used in modelling many 2014-02-11 56699.88 +8 500 FRODOSpec SNe Ia including 2003du (Tanaka et al. 2011), 2004eo (Mazzali 2014-02-13 56701.877 +10 500 FRODOSpec et al. 2008) and 2011fe (Mazzali, Sullivan & Hachinger 2014). a We present photometric and spectroscopic data taken with the Notes: Observation in MJD date. bRelative to B-band maximum. Liverpool Telescope (LT) and Isaac Newton Telescope (INT) for SN 2014J, and then apply the aforementioned modelling techniques Palma. Photometric observations were obtained using IO:O, an op- with the aim of inferring the ejecta properties on the SN. SN 2014J is tical imaging camera which has a field of view (FOV) of 10 arcmin2. a spectroscopically normal SN Ia, which has high-velocity features. The photometric images were acquired in three pass-bands (SDSS It is of particular interest as it is the closest SN Ia in at least the g r i). Spectra were obtained using FRODOSpec, a multipurpose last 28 yr, and possibly the closest in the last 410 yr (Foley et al. integral-field unit spectrograph, at 10 epochs from −8to+10 d 2014). As technology has dramatically improved in this time, this relative to B-band maximum. FRODOSpec consists of blue and red SN gives us a unique opportunity to intensely observe, analyse and arms which cover 3900–5700 Å and 5800–9400 Å, respectively. model a SN Ia with modern technology. SN 2014J is located at The IO:O pipeline carries out basic reduction of photometric = = RA 9:55:42 Dec. 69:40:26.0 (J2000), and is in M82 which is data; this consists of bias subtraction, trimming of the overscan ± at a distance of 3.77 0.66 Mpc. This value was obtained from regions and flat fielding. FRODOSpec has two independent reduc- 1 the mean distance from NED, which used two methods PNLF tion pipelines. The first one, L1, performs bias subtraction, overscan (Planetary Nebula Luminosity Function) and TRGB (Tip of the trimming and CCD flat fielding, whereas the second one, L2, is spe- Red Giant Branch) to derive it. However, it should be noted that cific to FRODOSpec (Barnsley, Smith & Steele 2012). It produces Foley et al. (2014) derive a distance of 3.3 Mpc. M82 is known sky-subtracted row-stacked spectra, which were used in this paper. for having a large amount of star formation; hence, it has a large No host galaxy subtraction has been performed when analysing the amount of dust (Hutton et al. 2014). Because of this, SN 2014J spectra, as no images were available. However, the sky subtrac- is highly and unusually reddened. It does not follow the average tion routine in the FRODOSpec pipeline removes most of the flux = galactic reddening law of RV 3.1. Detailed investigations into the from the M82, meaning that any flux which has not been subtracted host galaxy extinction were produced by Amanullah et al. (2014) should be negligible. and Foley et al. (2014). With our modelling approach, we optimize We also have two spectra obtained using the 2.5 m INT, lo- the published extinction values which produce the best fits of SN cated at ORM on La Palma. The Intermediate Dispersion Spectro- 2014J from the spectra. graph (IDS), a long-slit spectrograph on the INT, was used with The paper starts with a report of the observations (Section 2), the R1200Y grating and RED+2 camera. The INT observations − + which includes 12 spectra taken from 8to 10 d. In the following were made with the slit at the parallactic angle. Table 1 is a log of section (Section 3), we discuss the aperture photometry, including spectroscopic observations of SN 2014J. the SDSS g r i light curves. In Section 4, we present the fully reduced and calibrated spectra. The next section (Section 5) discusses the MC radiative transfer technique. In Section 6, the modelled 3 PHOTOMETRY spectra of 10 early-time epochs are presented and discussed. After- 2 Photometric reduction was carried out using the IRAF package wards (Section 7), we discuss the abundance stratification we have DAOPHOT. Instrumental magnitudes of the SN and stars within the inferred. Finally, the results are summarized and conclusions are field were obtained. In order to produce the calibrated magnitudes drawn from them (Section 8). of the SN, the colour terms and zero-points were required. For the g r and i filters, the colour terms were obtained by using the stan- 2 OBSERVATIONS dard star images, taken on the same night as the observations, and comparing their instrumental magnitudes to the APASS catalogue SN 2014J was discovered on 2014/01/21.810 by S. J. Fossey at magnitudes (Henden et al. 2009). IO:O typically requires an expo- the UCL observatory (Fossey et al. 2014). The LT carried out de- sure time of over 10 s for good photometry, due to the amount of tailed spectroscopic and photometric observations, starting from time it takes for the shutter to open and close. We typically have an 2014 January 22. The LT is a 2.0 metre fully robotic telescope located at Observatorio del Roque de los Muchachos (ORM) on La 2 IRAF is distributed by the National Optical Astronomy Observatories, which are operated by the Association of Universities for Research in Astronomy, 1 NASA/IPAC Extragalactic Database (NED). Inc., under cooperative agreement with the National Science Foundation.

MNRAS 445, 4424–4434 (2014) 4426 C.Ashalletal. Downloaded from https://academic.oup.com/mnras/article-abstract/445/4/4424/1078876 by Queen's University of Belfast user on 16 November 2018

Figure 2. SIFTO SN 2014J light-curve ﬁts.

2014). A large host galaxy reddening of SN 2014J was inferred by Amanullah et al. (2014) and Foley et al. (2014). The values obtained for the colour excess using the CCM law (Cardelli, Clayton & Mathis 1989) by Foley et al. (2014) were E(B − V) = 1.24 ± 0.1 mag and RV = 1.44 ± 0.06. However, Foley et al. (2014) state that the best solution to the extinction of SN 2014J is using a two-component cir- cumstellar scattering and dust reddening model. Amanullah et al. Figure 1. Aperture photometry light curves, produced using LT data. SDSS (2014) use photometric comparisons to SN 2011fe to obtain the g r i light curves are presented. The errors on the photometry appear to be best-fit for the reddening values. Using the FTZ reddening law constant; this is due to the catalogue error on the zero-point star dominating. (Fitzpatrick 1999) between −5and+35 d relative to B-band maximum it was found that E(B − V) = 1.29 ± 0.02 mag and an RV of 2 1.3 ± 0.1 with a χ /dof = 3.3. It should be noted that the RV value is exposure time of 2 s in our photometric images. Therefore, stars at much lower than the typical Galactic average of RV = 3.1. We use a the edge of the FOV will be exposed for a shorter period of time host galaxy extinction of E(B − V) = 1.2 with RV = 1.38; details of than ones in the centre of the field. To overcome this, the zero-points why we use these values can be found in Section 6. The attenuation were calculated using a single star, RA = 9:55:35 Dec = 69:38:55 of the flux in B and V is similar with the values of Amanullah et al. (J2000), close to SN 2014J. The main source of error in the photom- (2014) and Foley et al. (2014). etry was calculating the zero-points; if the calibration star was satu- Foley et al. (2014) and Marion et al. (2014) have published pho- rated the magnitude was ignored. As we do not have pre-explosion tometric and spectroscopic observations of SN 2014J. They report images of M82 using the LT, we have not subtracted the host galaxy m15(B) to be 1.01–1.08 mag when corrected for host galaxy ex- from the photometry. To calculate how much this host galaxy flux tinction, and 1.11 ± 0.02 mag when not corrected for host galaxy affects our photometry, we used SDSS images of M82. We carried extinction, respectively. We analysed the LT SDSS g-andr-band out aperture photometry, using the same aperture size as the LT photometry with the SIFTO (Conley et al. 2008) light-curve fitter to images, on the location of the SN pre-explosion and a standard star obtain the stretch, V-band maximum and tBmax (see Fig. 2). Us- in the FOV. From this, we found that the highest value the photom- ing the g and r bands, we obtained a stretch of 1.083 ± 0.06, etry can deviate by, owing to the missing host galaxy subtraction, is which corresponds to a m15(B) = 0.88 ± 0.08 using the rela- 2.35 per cent mag in the i band, 1 per cent mag in the r band and tion from Conley et al. (2008). However, using only the r-band 0.5 per cent mag in the g band. light curve produces a stretch of 1.035 ± 0.08 and therefore a SN 2014J was observed at a daily cadence from MJD 56680.20 m15(B) of 0.95 ± 0.12. When corrected for host galaxy extinc- to 56743.84. g r i light curves were produced (Fig. 1). There was a tion, E(B − V) = 1.2 (RV = 1.38), using the relation from Phillips gap in observations between 2014 February 13 and February 21, due et al. (1999), the B-band decline rate is found to be 1.00 ± 0.06 or to poor weather conditions on La Palma. g-band maximum occurred 1.07 ± 0.08 for g and r bands and r band, respectively. These de- on February 03, which is consistent with the maximum in the B pass- cline rates are consistent with those of Foley et al. (2014). However, bands (Tsvetkov et al. 2014). We found that g-band maximum was the correction of Phillips et al. (1999) for obtaining ‘reddening- 11.42 ± 0.09 mag, r -band maximum was 10.16 ± 0.03 mag and free’ m15 values from the observed values was derived assuming i -band maximum was 10.20 ± 0.05 mag. The distance modulus an RV = 3.1. To check the sensitivity of this correction to RV,we we use for M82 was taken from NED, μ = 27.86 mag. There- carried out our own calculation of the effect of dust reddening on fore, at maximum, SN 2014J had an absolute g -band magnitude m15(B)forRV = 3.1 and V1.4 using the published optical spec- of −16.44 mag, an r-band magnitude of −17.68 mag and an i-band trophotometry of SN 2011fe (Pereira et al. 2013)andtheHsiao magnitude of −17.66 mag, before correction for extinction. SN Ia spectral template (Hsiao et al. 2007). The spectra were red- The foreground Galactic extinction of SN 2014J is dened for values of E(B − V) = 0.0–2.0 using the IRAF DEREDDEN E(B − V) = 0.05 mag (RV = 3.1; Amanullah et al. 2014; Foley et al. task, which implements the CCM law. Synthetic magnitudes were

MNRAS 445, 4424–4434 (2014) SN 2014J models 4427 Downloaded from https://academic.oup.com/mnras/article-abstract/445/4/4424/1078876 by Queen's University of Belfast user on 16 November 2018

Figure 3. Dependence of m15(B)onE(B − V) for different values of RV, using SN 2011fe spectrophotometry and the Hsiao template. Figure 4. An overview of the features of SN 2014J. This spectrum has been dereddened and was taken at B-band maximum. calculated using the Bessell (1990) B pass-band, and the B-band decline rate was measured for each value of E(B − V). Fig. 3 shows the and the errors on the velocity. The velocity gradient from the Si II results, including a comparison with the approximate relation given 6355 Å line between +0and+9 d relative to B-band maximum is by Phillips et al. (1999). As is seen, the effect on m (B) of chang- 15 58.8 km s−1 day−1. This makes SN 2014J a low-velocity gradient ing the value of R from 3.1 to 1.4 is small. For E(B − V) = 1.2 and V SN Ia (Benetti et al. 2005). It can be seen in the LT post-maximum R = 1.38 application of these calculations to our SIFTO-measured V spectra at +6, +8and+10 d that P-Cygni re-emission redwards of decline rates gives m15(B) = 0.98 ± 0.08 or 1.05 ± 0.12, for the the Si II 6355 Å absorption has a flat top, which is also visible in the g and r bands and r band, respectively, using the 2011fe spectra, HST data at +11.3d(Foleyetal.2014). A full plot of a SN 2014J and m (B) = 0.95 ± 0.08 or 1.02 ± 0.12 using the Hsiao tem- 15 spectrum with the main absorption lines labelled is found in Fig. 4. plate. The value of the decline rate may increase when the full host By using the correlation between the ratio of the depth of the sil- galaxy subtraction can be carried out on the photometry. Foley et al. icon lines, 5972 and 6355 Å, and absolute magnitude and therefore (2014) state that the V-band maximum was 10.61 ± 0.05, whereas m (B) (Nugent et al. 1995) we have obtained a photometry- our fitting obtains this to be 10.66 ± 0.02. We obtain a t which 15 Bmax independent approximation for the m (B) of SN 2014J. We used is consistent with the values found by Foley et al. (2014). 15 the INT spectra from +1 d to do this, as the INT data cover the 5972 Å feature. We derive m15(B) = 1.11 ± 0.02, which is con- 4SPECTROSCOPY sistent with the values found by Marion et al. (2014) and Foley et al. (2014). Furthermore, it is possible to use the Equivalent Width The LT data spectroscopic reduction was done in two halves, corre- (EW)oftheSiII 5972 Å to estimate the m (B) (Hachinger, Maz- sponding to the blue and red arms of FRODOSpec. Each spectrum 15 zali & Benetti 2006). Using this method m (B) was found to be was manually searched through to select fibres which had signal. An 15 1.27 ± 0.15. appropriate top threshold was applied, to ensure cosmic rays were not affecting the spectrum. The signals from these fibres were com- bined; the spectra were formed using the ONEDSPEC IRAF package. 5 MODELLING TECHNIQUE The SN spectra were calibrated in flux using spectra of Feige34. The accuracy of the flux calibration process was confirmed by a By observing spectra at frequent intervals it is possible to study successful calibration of the star back on to itself. This was done by the detailed properties of SNe as a function of depth, as the photo- running the calibration process on the observations of the standard sphere recedes into the material with time and more and more of star, and checking this against the IRAF data for the star. The INT the ejecta mass becomes visible. We express this by a photospheric data were reduced using the STARLINK software (Disney & Wallace velocity decreasing in time, as velocity can be used as a comoving 1982). The spectrophotometric standard used to reduce the INT data spatial coordinate. This is because high velocities correspond to was Feige66. large radii, due to the homologous expansion of SNe Ia ∼10 s after A full plot of a SN 2014J spectrum with the main absorption the explosion. This can be approximated by equation (1), where r v lines labelled is found in Fig. 4 and the early-time spectral evo- is the distance from the centre of the explosion, ph is the velocity lution of SN 2014J is plotted in Fig. 5. The spectra have a very and texp is the time from explosion: strong Si II 6355 Å absorption line. They also have high-velocity r = v × t . (1) features at early epochs; the main high-velocity feature is exhibited ph exp by the Ca II IR triplet around ∼7900 Å. Furthermore, at even ear- The early-time spectra are of particular interest as they change lier times up to −12 d Goobar et al. (2014) have found a strong significantly over very small epochs, allowing us to infer abundance high-velocity feature of Si II 6355 Å. Fig. 5 demonstrates the ex- information about the outermost layers of the explosion. Carrying tent of the reddening, which means that the flux in the blue arm is on with the modelling up to post-maximum phases, the abundance extremely suppressed. The evolution of the Si II 6355 Å line can stratification is inferred for most of the ejecta except for the core be seen in Table 2. As expected, as the photosphere recedes, the layers, which become visible only in the nebular, late phase. This velocity of the ejecta decreases and hence the wavelength of the phase must be modelled separately (which is beyond the scope absorption line increases, although this may not always be apparent of this paper), owing to the non-local-thermodynamic-equilibrium over smaller intervals. This is due to the resolution of the spectra state of the plasma at late epochs.

MNRAS 445, 4424–4434 (2014) 4428 C.Ashalletal. Downloaded from https://academic.oup.com/mnras/article-abstract/445/4/4424/1078876 by Queen's University of Belfast user on 16 November 2018

Figure 5. All spectral observations of SN 2014J, LT and INT. The time is given relative to rest-frame B-band maximum. The spectra have not been corrected for reddening. There were no data collected between 5700–5800 Å for the LT observations and the two atmospheric absorption lines have been removed. All of the plots have been offset by an arbitrary amount for the purpose of presentation.

A useful way to model early-time spectra, which still obtains of radioactive heating below the photosphere. Furthermore, since information from observational data, is to use MC radiative transfer all of the important absorption lines for the abundances are in the codes. Using this approach over a number of high and frequent UV/optical this excess flux in the IR has little effect on the results. epochs leads to an accurate abundance distribution of the SN. The The exact 56Ni distribution below the photosphere is therefore irrel- MC radiative transfer code here is based on code originally written evant to the calculation, it is only important that the bulk of the 56Ni by Abbott & Lucy (1985) in relation to stellar winds. This was of the SN is below the photosphere. Therefore, the code can only be adapted by Mazzali & Lucy (1993) for early-time SN spectra, and used up to ∼14dpastB-band maximum, after this the assumptions further improved by Lucy (1999) and then Mazzali (2000). of the code are weak. We obtain optimally fitting synthetic spectra for the photospheric The code has a selection of input parameters: the bolometric lu- phase as Stehle et al. (2005) did with their abundance tomography minosity, ejecta velocity and chemical abundance stratification. The approach. This technique has been used to model a number of SNe, typical process of stratification tomography consists of modelling including SN 2002bo (Stehle et al. 2005), SN 2010jn (Hachinger, the earliest spectrum then moving inwards until the last photospheric Mazzali & Sullivan 2013) and SN 2011fe (Mazzali et al. 2014). spectrum is reached. The initial abundances are set by an educated The code assumes 1D spherically symmetric ejecta and a grey approximation from a typical explosion model. The luminosity is photosphere from which radiation is emitted, and calculates the then iterated until it matches the flux of the observed spectra. This is interaction of photons within the expanding SN ejecta. The code followed by the iteration of the velocity to fit the observed blueshifts simulates the propagation of emitted photons by considering them of the P-Cygni features, which in turn fits the position of the lines. as photon packets, which can undergo Thomson scattering and line Finally, the abundances are changed to ensure that the absorption absorption. At the photosphere, below which the ejecta are assumed line strengths are all modelled. The abundance for a new inner to be optically thick, the outward flowing radiation is assumed to be layer is calculated. The abundance tomography technique ensures from a blackbody. This assumption can cause excess flux in the IR that this spectrum shows absorption from the outer layers but also and red side of the spectra at later epochs. However, it has the ad- additional layers which are above the new photosphere and below vantage that the abundances can be derived without the knowledge the old one. This modelling process is repeated for each spectrum

MNRAS 445, 4424–4434 (2014) SN 2014J models 4429

Table 2. Log of the Si II 6355 Å absorption line Table 3. Input parameters and calculated converged velocities. temperature.

a b Epoch MJDa Phaseb Velocityc Epoch Epoch Velocity Bol. lum. Temp. − (d) (d) (km s 1) tB texp vph TBB (d) (d) (km s−1) log(L)(K) 2014-01-26 56684.12 −8 12 973 2014-01-27 56685.12 −7 12 643 −8 12 12 300 9.295 10 200 Downloaded from https://academic.oup.com/mnras/article-abstract/445/4/4424/1078876 by Queen's University of Belfast user on 16 November 2018 2014-02-03 56692.01 +0 11 758 −7 13 11 990 9.408 10 100 2014-02-04 56692.9 +1 11 697 +0 20 9480 9.455 9200 2014-02-05 56693.89 +2 11 686 +1 21 8970 9.46 9000 2014-02-06 56694.96 +3 11 552 +2 22 8440 9.43 8900 2014-02-07 56695.89 +4 11 640 +3 23 7930 9.39 8800 2014-02-08 56696.89 +5 11 412 +4 24 7480 9.355 8600 2014-02-09 56697.97 +6 11 529 +6 26 6500 9.32 8400 2014-02-11 56709.87 +8 11 408 +8 28 5450 9.250 8100 2014-02-13 56701.877 +10 11 228 +10 30 4400 9.22 7800 a Notes: aModified Julian Day number. Notes: Relative to B-band maximum. b bRelative to B-band maximum (days). Days after the explosion. c Velocity of the Si II 6355 Å absorption line. bolometric rise time of ∼20 d. The main input parameters can be and previous abundances in outer layers may need to be changed in found in Table 3. order to fit the later spectra, in which case the entire fitting process The main limitations of the analysis concern the high-velocity is repeated. outer layers of the ejecta, because we do not have early-time spectra, Choosing a reasonable density profile is important in producing the extinction values, the distance to SN 2014J and the lack of a physically meaningful and well-fitting model. The density distri- UV data. Although we may expect uncertainty to be of the order bution chosen for SN 2014J is the W7 model (Nomoto et al. 1984). of ∼10 per cent, the biggest uncertainty is due to the lack of early This is a fast deflagration explosion of a Chandrasekhar mass C+O UV data. The UV data will be modelled in a follow-up paper using WD. The deflagration wave synthesizes 0.5–0.6 M 56Ni in the the HST data. inner layer of the star, which is enough to power the light curve of This section will analyse each epoch of the models and observa- the SN (Nomoto et al. 1984). The W7 model was selected as SN tions. In the models we parametrize the iron-group content in terms 2004eo, SN 2003du and SN 2002bo can all be reasonably modelled of two quantities, Fe and 56Ni at the time t = 0. This gives us the with this density profile. Therefore, we can see if there is conti- abundance of Fe, 56Co and 56Ni at any point of time, assuming that nuity in results. A higher density can lead to enhanced absorption the abundances of directly synthesized Co and Ni are negligible. lines, therefore it is important to select the appropriate one for the Therefore, it should be noted that any Fe discussed in this section explosion. The most marked effects of the density on photospheric is stable iron. spectra occur at the earliest epochs, when the density of the outermost layers strongly modulates, e.g. the UV flux (e.g. Mazzali et al. − 2014). Therefore, moderate deviations in the outer density profile 6.1 Day 8 will not affect the results obtained from the regions explored in this The first spectrum was observed 8 d before B-band maximum, paper, as we begin with the modelling at an epoch of −8 d relative texp = 12 d (see Fig. 6). As this spectrum is before B-band maxi- to B-band maximum. Still, in follow-up papers we will test density −1 mum it has a high photospheric velocity, vph = 12 300 km s ,the profiles of DD and sub-Chandrasekhar-mass models. effective temperature is 10 200 K and the bolometric luminosity is 109.295 L. The strongest photospheric absorption lines, which we have indicated in Fig. 4 (Section 4), are dominant from the 6 MODELS beginning of the time series to at least a week after maximum. 10 epochs of LT data have been modelled using the abundance There are strong Si II 6355 Å, S II 5454 Å and 5640 Å features. tomography technique. All of the spectra have been dereddened Due to this, the model requires a photospheric abundance of 12 per using the CCM law. In this paper, we use a host galaxy extinction cent Si and 9 per cent S by mass, with 13 per cent Mg to produce of E(B − V) = 1.2 mag (RV = 1.38). These values are where our the Mg II 4481 Å line. The Mg II 4481 Å line is the dominant line in model produce the best fits, and are also consistent with the values the 4300 Å feature. The O abundance is 60 per cent and the Ca is of Amanullah et al. (2014) and Foley et al. (2014). The MC radiative 1 per cent. There are strong Ca II features in this modelled spectrum, transfer code is successful at modelling a variety of SN; therefore, including a large absorption line in the H&K feature in the near UV the models of SN 2014J are a good indication that the derived (3934, 3968 Å). This absorption line is produced by our models in a reddening values are correct. One-shell models were produced with strength consistent with the HST data (Foley et al. 2014). Although different values of E(B-V), and it was found that if this value was we do not cover this UV Ca line in our observations, we do have increased the input luminosity of the model had to be increased. the Ca II IR triplet from which we can infer the Ca abundances. The This meant there was too much flux in the model and it peaked in photospheric Ca II IR triplet is modelled successfully, but there is the UV rather than the optical. no attempt to model the high-velocity feature. An iron abundance The time since explosion is one of the input parameters needed of 4 per cent is required to produce the Fe III 5150 Å absorption for the modelling process; therefore, the rise time of SN 2014J is line. The 56Ni abundance is at 0 per cent, because the ejecta at this needed. There are pre-discovery images of SN 2014J, in which it epoch are still in the high-velocity outer layers of the photosphere. first appears at Jan. 14.75 UT (Zheng et al. 2014) which gives it a Ti+Cr are at a photospheric abundance of 0.3 per cent by mass.

MNRAS 445, 4424–4434 (2014) 4430 C.Ashalletal. Downloaded from https://academic.oup.com/mnras/article-abstract/445/4/4424/1078876 by Queen's University of Belfast user on 16 November 2018

Figure 6. Modelled (blue) and observed spectra (green) of SN 2014J. Plot (a) is from −4 d, (b) is from −3 d, (c) is +0 d maximum, (d) is +1 d and (e) is +2 d. All of the dates are relative to B-band maximum. All spectra have been dereddened. The red spectrum at +1 d is from the INT.

There is a low abundance of C at this high velocity as there are no changes from the previous epoch are that Si has increased to a C spectral features in the optical data, suggesting that all the carbon photospheric mass abundance of 15 per cent and S to 10 per cent. may be in earlier time spectra, although Marion et al. (2014)doﬁnd The 56Ni has also increased to 3 per cent. The ∼4200 Å feature is C I at 1.0693 µm. However, if we add C into the models we ﬁnd predominately Mg II 4481 Å with smaller contributions from Fe III this produces deeper and wider C absorption lines. Therefore, the C 4419 Å. The Mg abundance is 10 per cent and the O abundance has found in the NIR could be due to a very small abundance. There is decreased to 56 per cent. The Si III 4550 Å line is successfully mod- a small absorption line in the optical observed at ∼6200 Å, where elled, and this is the last FRODOSpec spectrum with a prominent one would expect C, however this line is too narrow and not a C Si III feature. feature. The model at this epoch produces a particularly good match for the Fe II/Mg II 4200 Å, S II 5640 Å and Si II 6355 Å features. 6.3 Day +0

− The third spectrum was taken on the night of B-band maximum, 6.2 Day 7 9.455 texp = 20 d (refer to Fig. 6). The luminosity is 10 L and −1 The second spectrum was taken −7 d from B-band maximum, photospheric velocity vph = 9480 km s . The Si and S abundances 9.408 texp = 13 d. The luminosity at this epoch is 10 L and have increased relative to the previous epoch to photospheric mass −1 56 vph = 11990 km s . There is very little variation between the ﬁrst abundances of 17 and 12 per cent, respectively, and the Ni has two epochs as they are only taken 1 d apart. The main chemical stayed constant at 3 per cent. 56Fe has also increased by 2–6 per

MNRAS 445, 4424–4434 (2014) SN 2014J models 4431 cent. Conversely, the Mg and O abundances have decreased to 5 and Si and S have also decreased to an abundance of 4 and 2 per cent 55 per cent, respectively. There is now a notable excess in ﬂux in the by mass, respectively. The Fe II 4420 Å and Si II 6355 Å lines are red side of the spectrum due to the blackbody approximation. The modelled successfully. The wide deep absorption line at 8200 Å is B-band modelled absolute unreddened magnitude is −18.79. At this the Ca II IR triplet absorption line, the calcium has been reduced epoch, the modelled Si III 4550 Å absorption line is not as strong to a 1 per cent by mass to decrease the strength of this line. At as the previous epochs. Furthermore, in the previous spectrum the this epoch 56Ni begins to dominate and is at 88 per cent. Given the 56 4800 Å feature which have dominant Si II 5056 Å and Fe II 5169 Å sudden jump in the Ni abundance, we can give the constraint that Downloaded from https://academic.oup.com/mnras/article-abstract/445/4/4424/1078876 by Queen's University of Belfast user on 16 November 2018 absorption lines are merged into one, whereas at this epoch they the 56Ni distribution will extend to ∼8000 km s−1. We thus predict have two distinct minima. Ca absorption is now more prominent that the characteristic line width of the iron emission in the nebula −1 in both the model and the observations, and is seen in the Ca II IR spectra will be of the order of ∼8000 km s (Mazzali et al. 1998). triplet at ∼8200 Å.

6.4 Day +1 6.7 Day +4 The next spectrum in the series was observed on +1 d relative to The spectrum from +4 d has a luminosity of 109.345 L and photo- 9.460 −1 B-band maximum, texp = 21 d. It has a luminosity of 10 L spheric velocity of vph = 7480 km s . There is a similar discrepancy −1 and a photospheric velocity of vph = 8970 km s . This spectrum between the INT and LT data as in the +1 d plot; however, once is at maximum bolometric luminosity, which is consistent with it again this does not affect the abundances we obtain from our fits. being between B-band and V-band maximum. Although the INT From this epoch, the absorption lines are beginning to be stronger and LT spectra differ slightly, predominantly in the red side of the than the observed ones. For example, the modelled Fe/Mg 4300 Å spectrum, the main absorption lines in the blue side of the optical are and Fe 4800 Å features, which are produced by dominant Fe III similar between the spectra. Due to this, the abundances obtained in 4419 Å and Fe II 5018 Å absorption lines. However, to refine this our analysis would not change if we were to model just the INT or fit requires the abundances of these lines at high-velocity epochs LT spectra. The abundances we derive from this spectrum are very to be reduced, affecting the early-time spectra. Therefore, we sug- similar to the previous epoch; the Si decreases to 15 per cent, the gest that the excess absorption could be due to too much mass at 56 −1 56 S to 10 per cent the O to 57 per cent. Ni has increased to 11 per vph = 7480 km s . Part of the excess in strength of the Fe lines cent and 56Fe is constant at 6 per cent. The effective temperature could also be due to the decay of 56Ni. Due to this excess strength at this epoch is 9000 K, which is 200 K lower than the previous in Fe, its photospheric abundance is now at 0.1 per cent. The Si and spectrum. The Mg abundance has decreased to 0 per cent. The S abundances have decreased to 0 per cent, and the 56Ni is at 99 per S II 5454 Å feature is not as strong in the model as the observed cent. The effective temperature at this epoch is 8600 K. spectrum; however, increasing S abundance would enhance the S II 5640 Å line. There is excess strength of O I and Mg II at 7773 and 7896 Å which could be an indication that there is excess mass at 6.8 Day +6 this velocity. This feature occurs in most of the epochs in the model, and is more dominant in late-time spectra. The eighth spectrum was taken 26 d after explosion. It has a lumi- 9.32 −1 nosity of 10 L, a photospheric velocity of vph = 6500 km s and an effective temperature of 8400 K. The 56Ni photospheric 6.5 Day +2 abundance is at 99 per cent. The Fe II 4340 Å, S II 5606 Å and The fifth spectrum was observed on texp = 22 d (see Fig. 6). Its lumi- Fe III 4419 Å absorption lines are much deeper in the model than 9.430 −1 nosity is 10 L and photospheric velocity is vph = 8440 km s . the observations. The unreddened modelled B and V magnitudes are The effective temperature is 8900 K. The abundances at this epoch 9.5 and 9.1 mag, respectively. are very similar to the previous one, except for 56Ni which begins to increase dramatically to 36 per cent, S which decreases to 4 per cent; Si and O also decrease to 10 and 43 per cent, respectively. At 6.9 Day +8 this epoch, Ca stays constant at 2 per cent. The B-band unreddened modelled absolute magnitude of this spectra is −18.85. At this The next spectrum has a luminosity of 109.250 L and photospheric −1 56 epoch the ∼4200 Å feature is still dominated by the Fe III 4419 Å velocity is vph = 5950 km s .The Ni abundance has stayed con- and Mg II 4481 Å lines. stant. The difference between the model and observation begins to differ even more, as shown by the Fe II 4549 Å line in the 4800 Å feature. At this epoch, there is a significant amount of 56Ni above 6.6 Day +3 the photosphere. Therefore, it is not unexpected that the difference The next spectrum was observed on texp = 23 d (refer to Fig. 7). The between the models and observed spectra begins to increase. luminosity at this epoch is 109.390 L, the photospheric velocity is −1 vph = 7930 km s , the effective temperature is 7900 K and the B- band modelled absolute magnitude is −18.622. The modelled S II 6.10 Day +10 5640 Å feature is stronger than the observed one; this is a problem which consistently occurs in the model. To make the S II the same The final epoch that was modelled is from +10 d from B-band strength as the observed one it would require the abundance of S to maximum. It has a luminosity of 109.22 L, a photospheric velocity −1 be reduced in the early-epoch models. However, we have chosen to of vph = 5450 km s and an effective temperature of 7800 K. At fit the early-time spectra rather than the late-time ones, as these can this epoch there is no Si or S abundance, and the 56Ni abundance is lead us to more information about the high-velocity abundances. At the most dominant, at almost 100 per cent. The Ca abundance has this epoch, the O has a photospheric abundance of 0 per cent. The now decreased to 0 per cent.

MNRAS 445, 4424–4434 (2014) 4432 C.Ashalletal. Downloaded from https://academic.oup.com/mnras/article-abstract/445/4/4424/1078876 by Queen's University of Belfast user on 16 November 2018

Figure 7. Modelled (blue) and observed spectra (green) of SN 2014J. Plot (a) is from +3 d, (b) is from +4 d, (c) is +5d,(d)is+6 d and (e) is +8d.Allof the dates are relative to B-band maximum. All spectra have been dereddened. The red spectrum of +4 d is from the INT.

7 ABUNDANCE STRATIFICATION At high velocities, between 8440–14800 km s−1, there is a large We produce an abundance tomography distribution plot for the oxygen abundance which starts at 78 per cent. The Si distribution starts at 10 per cent due to the strong Si II 6355 Å feature, and photospheric layers of the ejecta, see Fig. 8. This plot demonstrates −1 − + it increases to 17 per cent at 9480 km s before it decreases to how the abundances in the early times, from 8to 10 d, of SN −1 2014J develops. We cannot definitively confirm the abundances 0 per cent at 7480 km s . Sulphur also follows a similar distri- of the outermost layers of the ejecta as we do not have spectra bution with respect to velocity, although it always has a smaller before −8 d. The abundances in the outer shell are slightly unusual, abundance than Si. The sulphur distribution starts at 3.5 per cent in that there is no carbon abundance. The other notable abundance before peaking at 12 per cent. The basic abundance evolution of the is Fe which begins at 0.1 per cent and rises to 6 per cent. We have ejecta involves O dominating followed by the Intermediate Mass attempted to increase this, but doing so dramatically strengthens the Elements (IME) and then by the heavy elements. In the abun- Fe 4800 Å feature at all epochs. Therefore, the initial abundances of dance distribution plot, Fig. 8, the Fe starts at 0.1 per cent and the outer shell, which has a velocity of 14 800 km s−1, are Si 10 per rises to 6 per cent. In Fig. 8, the IME elements are significant at cent, O 78 per cent, S 4 per cent, 56Ni 0 per cent, Mg 7 per cent, C high velocities. From this, it can be inferred that the lighter ele- 0.0 per cent and 56Fe 0.1 per cent, with heavier elements making up ments may be at even higher velocities. Therefore, earlier spec- the remaining abundance. tra are needed to gain information about these lighter elements.

MNRAS 445, 4424–4434 (2014) SN 2014J models 4433

or 0.95 ± 0.12 using the g and r band or r band, respectively, and by ﬁtting them through SIFTO. When correcting for reddening, this produced values of m15(B) = 0.98 ± 0.08 or 1.05 ± 0.12, respectively, using the SN 2011fe spectra, and m15(B) = 0.95 ± 0.08 or 1.02 ± 0.12 for the Hsiao template. This result is consistent with those of Foley et al. (2014). We obtain a Vmax of 10.66 ± 0.02.

SN 2014J is a highly reddened SN Ia which does not follow the Downloaded from https://academic.oup.com/mnras/article-abstract/445/4/4424/1078876 by Queen's University of Belfast user on 16 November 2018 conventional Galactic reddening law (RV = 3.1). We adopt the CCM law with a foreground galactic value of E(B − V) = 0.05 (RV = 3.1) and host galaxy extinction of E(B − V) = 1.2 (RV = 1.38). SN 2011fe and SN 2014J were found to have comparable Ni masses, 0.4–0.7 and 0.47–0.72 M, respectively, when modelled using the same MC radiative transfer code. However, due to their difference in uncorrected decline rate we could expect these values to change. This could be revealed with the use of the UV data and the nebular spectra. Figure 8. The abundance distribution of SN 2014J. The Ni/Co/Fe abun- The spectra were modelled with the abundance tomography tech- 56 56 dances are given in terms of mass fractions of Ni ( Ni0) and stable Fe nique of Stehle et al. (2005), simulating spectrum formation with = (Fe0)att 0. In the spectral models no stable Ni or Co and no radioactive an MC radiative transfer code. The density profile used for this was Fe is assumed to be present. W7. In follow-up papers, we will change the density structure in an attempt to refine the model. 56 As expected from a normal SN Ia explosion Ni dominates the The spectra were modelled at 10 epochs, before and after B-band −1 abundance. This happens between 8440–7930 km s , where the maximum, inferring a best-fitting abundance stratification. As one 56 Ni goes from a photospheric mass fraction of 36 to 83 per cent. The would expect, at higher velocities (12 400 km s−1) there is a large Ti+Cr abundances stay at a constant level throughout the whole ex- abundance of oxygen. As the photosphere recedes (8440 km s−1) −1 plosion at 0.3 per cent. Calcium starts at 2 per cent at 12 300 km s the IME elements dominate, Si and S. Then at low velocities the −1 and it decreases to 1 per cent at 5450 km s . radioactive Ni dominates below 8000 km s−1. This leads to the pre- The integrated abundances of the most important elements in diction that the characteristic line width of the iron emission line −1 the photospheric ejecta, which is at a velocity above 4400 km s , in the nebular spectra will be of the order of 8000 km s−1 (Mazzali are M(Mg) = 0.07 M,M(Fe)= 0.03 M,M(O)= 0.40 M, et al. 1998). 56 M(S) = 0.058 M,M(Si)= 0.09 M and M( Ni) = 0.47 M. Synthetic spectra reproduce the spectral evolution of SN 56 When the nebular phase spectra is available, the Ni could in- 2014J, and the final integrated abundances of SN 2014J are crease to a total integrated abundance of 0.72 M. The final total M(Mg) = 0.07 M,M(Fe)= 0.03 M,M(O)= 0.40 M, integrated abundances can be confirmed when late-time spectra of M(S) = 0.05 M, M(Si) = 0.09 M and M(56Ni) = 0.47–0.72 M. SN 2014J are observed. Due to ground-based telescopes not being Our results are consistent with the current understanding of SN Ia able to observe the UV part of the spectra, the iron-group element reddening and early-time abundance distribution. The observation abundances of SN 2014J we have given here may show some 20 per and modelling in this paper is of particular significance because cent uncertainty (of the values given), which nebular modelling will of the close proximity of SN 2014J. Furthermore, SN 2014J is the allow us to approve upon. typical example of a normal SN Ia, making our models a good basis SN 2011fe and SN 2014J are photometrically similar. SN 2011fe for studying further objects. has been used to determine the extinction of SN 2014J (Amanullah et al. 2014), and it has also been modelled using the same MC radiative transfer code (Mazzali et al. 2014). The total 56Ni abundance ACKNOWLEDGEMENTS of SN 2011fe (Mazzali et al. 2014) and SN 2014J are very similar SH is funded from a Minerva ARCHES award of the Ger- (0.4–0.7 M and 0.47–0.72 M, respectively; the large range is man Ministry of Education and Research (BMBF). We have due to not knowing abundance distribution in the nebula phase). used data from the NASA / IPAC Extragalactic Database (NED, Furthermore, the abundances in the photospheric region of the SN http://nedwww.ipac.caltech.edu, operated by the Jet Propulsion 2011fe ejecta are remarkably similar to those of SN 2014J. Chang- Laboratory, California Institute of Technology, under contract with ing the density profile in the models is not likely to qualitatively the National Aeronautics and Space Administration). For data han- affect the abundance pattern in the regions of the ejecta explored by dling, we have made use of various software packages (as mentioned the spectra. in the text) including IRAF (Image Reduction and Analysis Facility). IRAF (http://iraf.noao.edu) is an astronomical data reduction soft- 8 SUMMARY ware distributed by the National Optical Astronomy Observatory (NOAO, operated by AURA, Inc., under contract with the National We have presented photometric and spectroscopic observation of the Science Foundation). closest SN Ia in at least the last 28 yr, SN 2014J. The observations were obtained with the LT and INT. We have presented SDSS g r i light curves and a spectral time series evolution over 12 epochs REFERENCES − + from 8to 10 d, relative to B-band maximum. All of the spectra Abbott D. C., Lucy L. B., 1985, ApJ, 288, 679 were calibrated in flux and atmospheric absorption lines were re- Amanullah R. et al., 2014, ApJ, 788, L21 moved. The spectra show a very deep Si II 6355 Å line and a Ca high- Barnsley R. M., Smith R. J., Steele I. A., 2012, Astron. Nachr., 333, 101 velocity feature at ∼7900 Å. We obtain a m15(B) = 0.88 ± 0.08 Benetti S. et al., 2005, ApJ, 623, 1011

MNRAS 445, 4424–4434 (2014) 4434 C.Ashalletal.

Bessell M. S., 1990, PASP, 102, 1181 Mazzali P. A., Lucy L. B., 1993, A&A, 279, 447 Cardelli J. A., Clayton G. C., Mathis J. S., 1989, ApJ, 345, 245 Mazzali P. A., Cappellaro E., Danziger I. J., Turatto M., Benetti S., 1998, Conley A. et al., 2008, ApJ, 681, 482 ApJ, 499, L49 Disney M. J., Wallace P. T., 1982, QJRAS, 23, 485 Mazzali P. A., Sauer D. N., Pastorello A., Benetti S., Hillebrandt W., 2008, Fitzpatrick E. L., 1999, PASP, 111, 63 MNRAS, 386, 1897 Foley R. J., Fox O., McCully C., Phillips M., 2014, MNRAS, 443, 2887 Mazzali P. A., Sullivan M., Hachinger S., 2014, MNRAS, 439, 1959 Fossey J., Cooke B., Pollack G., Wilde M., Wright T., 2014, Cent. Bur. Nomoto K., Thielemann F.-K., Yokoi K., 1984, ApJ, 286, 644 Electron. Telegrams, 3792, 1 Nomoto K., Iwamoto K., Kishimoto N., 1997, Science, 276, 1378 Downloaded from https://academic.oup.com/mnras/article-abstract/445/4/4424/1078876 by Queen's University of Belfast user on 16 November 2018 Goobar A., Johansson J., Amanullah R., Cao Y., Perley D. A., Kasliwal M. Nugent P., Phillips M., Baron E., Branch D., Hauschildt P., 1995, ApJ, 455, M., Ferretti R., 2014, ApJ, 784, L12 L147 Hachinger S., Mazzali P. A., Benetti S., 2006, MNRAS, 370, 299 Pereira R., Thomas R. C., Aldering G., Antilogus P., Baltay C., 2013, A&A, Hachinger S., Mazzali P. A., Sullivan 2013, MNRAS, 429, 2228 554, A27 Henden A. A., Welch D. L., Terrell D., Levine S. E. 2009, AAS, 214, 0702 Perlmutter S., Aldering G., Goldhaber G., Knop R. A., Nugent P., Castro P. Hoyle F., Fowler W. A., 1960, ApJ, 132, 565 G., 1999, ApJ, 517, 565 Hsiao E. Y., Conley A., Howell D. A., Sullivan M., Pritchet C. J., Carlberg Phillips M. M., Lira P., Suntzeff N. B., Schommer R. A., Hamuy M., Maza R. G., Nugent P. E., Phillips M. M., 2007, ApJ, 663, 1187 J., 1999, ApJ, 118, 1766 Hutton S., Ferreras I., Wu K., Kuin P., Breeveld A., Yershov V., Cropper M., Riess A. G., 1998, ApJ, 116, 1009 Page M., 2014, MNRAS, 440, 150 Stehle M., Mazzali P. A., Benetti S., Hillebrandt W., 2005, MNRAS, 360, Iben I., Jr, Tutukov A. V., 1984, ApJS, 54, 335 1231 Iwamoto K., 1999, ApJS, 125, 439 Tanaka M., Mazzali P.A., Stanishev V.,Maurer I., Kerzendorf W.E., Nomoto Khokhlov A. M., 1991, in Danziger I. J., Kjaer K., eds, ESO Conf. Work- K., 2011, MNRAS, 410, 1725 shop Proc. Vol. 37, Delayed Detonation Model for Type ia Supernovae. Tsvetkov D. Y., Metlov V. G., Shugarov S. Y., Tarasova T. N., Pavlyuk N. European Southern Observatory, Garching, p. 161 N., 2014, preprint (arXiv: e-prints) Leibundgut B. et al., 1993, ApJ, 105, 301 Zheng W. et al., 2014, ApJ, 783, L24 Livne E., Arnett D., 1995, ApJ, 452, 62 Lucy L. B., 1999, A&A, 345, 211 Marion G. H. et al., 2014, preprint (arXiv: e-prints) Mazzali P. A., 2000, A&A, 363, 705 This paper has been typeset from a TEX/LATEX ﬁle prepared by the author.

MNRAS 445, 4424–4434 (2014)